Whisper Web brings the Whisper speech recognition model into your browser. You drag an audio or video file onto the page, and the transcription happens on your own device rather than on a remote server, so the file is never uploaded.
The site lists support for more than 99 languages and puts accuracy at around 98 percent. It is free to use and needs no installation, which makes it a practical choice for Pakistani students and journalists transcribing lectures, interviews or recordings.
